[OSGeo-Discuss] [Board] Request for a moratorium on AI/LLM use within OSGeo

Alex Kmoch allixender at gmail.com
Mon Nov 2 02:01:41 PST 2026


Hi,

I think Adam's proposed approach is not bad, but reducing this to a 
singular "the foundation doesn't care" is a bit harsh isn't it?

Just as a side info maybe for additional context: In European 
universities we have a similar discussion, but the stand-point is 
developing mostly towards that we rather should deal with AI/LLM usage 
more proactively. We cannot ignore it away. People gonna use it, and if 
you don't "recognize" AI/LLM usage in a pull request, but the user 
actually used AI to better understand and improve code usefully in 
manageable pieces, and gives it his own friendly message on the way, 
would you still reject it? Where do you draw the line? Why not instead 
considering building guidelines and framing expectations, and maybe have 
use OSGEO community think about constructive ways, not just outright 
banning?

Even has indicated he has a fundamental problem with AI/LLM in the 
world, whereas dealing with AI/LLM generated code (and maybe docs) 
within OSGEO I would say calls for a bit more nuance?

Alex


On 30.10.2026 09:29, Even Rouault via Discuss wrote:
> The lack of communication from the board on this critical topic raises 
> the question of whether it really cares about its projects, and if its 
> projects should really care about the foundation...
>
> Le 15/10/2026 à 20:59, Adam Steer via Discuss a écrit :
>> Hi all
>>
>> A concrete action: write a statement saying "LLM generated code pull 
>> requests and bug reports are not permitted in OSGeo projects. Neither 
>> are LLM generated documents, presentations, communications".
>>
>> Vote on it. Add it to the company constitution / project guidelines / 
>> FAQs / wherever relevant.
>>
>> As well as approaches already linked up thread, a broader model for a 
>> rationale can be found at [1]
>>
>> On this the board does not need to be burdened. If the intent of 
>> OSGeo is to support and stand for community led collaboration then 
>> the course of action is clear.
>>
>> [1] https://www.spatialised.net/no-ai/
>>
>> Regards,
>>
>> -- 
>> Dr Adam Steer
>> https://spatialised.net
>> https://mapstodon.space/@adamsteer
>>
>> _______________________________________________
>> Discuss mailing list
>> Discuss at lists.osgeo.org
>> https://lists.osgeo.org/mailman/listinfo/discuss
>
-- 
Geospatial tinkerer

WW: https://github.com/allixender/
FV: https://fosstodon.org/@allixender



More information about the Discuss mailing list